turicreate.evaluation.max_error

turicreate.evaluation.max_error(targets, predictions)

Compute the maximum absolute deviation between two SArrays.

Parameters:

targets : SArray[float or int]

An Sarray of ground truth target values.

predictions : SArray[float or int]

The prediction that corresponds to each target value. This vector must have the same length as targets.

Returns:

out : float

The maximum absolute deviation error between the two SArrays.

See also

rmse

Notes

The maximum absolute deviation between two vectors, x and y, is defined as:

$$\textrm{max error} = \max_{i \in 1,\ldots,N} \|x_i - y_i\|$$

Examples

>>> targets = turicreate.SArray([3.14, 0.1, 50, -2.5])
>>> predictions = turicreate.SArray([3.1, 0.5, 50.3, -5])
>>> turicreate.evaluation.max_error(targets, predictions)
2.5